Template Recommendation Method, Apparatus, Device, and Storage Medium

Abstract

The present disclosure provides a template recommendation method, apparatus, device, and storage medium, which includes first obtaining features of a multimedia material to be processed, and determining a similarity between the features corresponding to the multimedia material to be processed and the features corresponding to a candidate editing template, and determining the candidate editing template as a target editing template corresponding to the multimedia material to be processed if it is determined that the similarity satisfies a preset matching condition.

Description

Technical Field

[0001] [Cross - Reference to Related Applications] This disclosure claims the priority of a patent application filed on June 24, 2021, with the application title "Template Recommendation Method, Apparatus, Device, and Storage Medium" and the Chinese patent application number "202110703796.7", and all the contents of the Chinese patent application are incorporated into this disclosure by reference. This disclosure relates to the field of data processing, and particularly to a template recommendation method, apparatus, device, and storage medium.

Background Art

[0002] With the development of computer technology, users' requirements for processing multimedia materials are getting higher and higher. At the same time, users' willingness to actively participate in the processing of multimedia materials is also increasing.

[0003] However, since many users lack expertise in processing multimedia materials, editing templates for multimedia material processing have emerged accordingly. Users can actively select an editing template and multimedia materials, and can generate a multimedia work with one touch. Obviously, such a method lowers the threshold for processing multimedia materials, and such a simple and convenient way of generating multimedia works is also favored by many users.

[0004] In actual applications, the editable templates that users can select are generally templates with relatively high recent usage frequencies by users, that is, editable templates with relatively high popularity. However, editable templates with relatively high popularity are not necessarily applicable to the multimedia materials selected by users, which may lead to situations such as the failure to generate multimedia works, and the user experience is reduced.

Summary of the Invention

Problems to be Solved by the Invention

[0005] Therefore, how to recommend an editing template suitable for the multimedia material selected by the user and improve the user's satisfaction with the editing template recommendation is a technical problem to be solved currently.

Means for Solving the Problem

[0006] To solve or at least partially solve the above technical problem, the present disclosure provides a template recommendation method, apparatus, device, and storage medium.

[0007] In a first aspect, the present disclosure obtaining characteristics of the multimedia material to be processed, including content characteristics of the multimedia material to be processed and / or user attributes corresponding to the multimedia material to be processed; determining a similarity between the characteristics corresponding to the multimedia material to be processed and the characteristics corresponding to a candidate editing template, the characteristics corresponding to the candidate editing template being determined based on a historical multimedia work that is a multimedia work edited in an editing manner indicated by the candidate editing template; when it is determined that the similarity satisfies a preset matching condition, determining the candidate editing template as a target editing template corresponding to the multimedia material to be processed, the target editing template being used to instruct the multimedia material to be processed to be edited into a target multimedia work according to the editing manner; and providing a template recommendation method including the above.

[0008] In one selectable embodiment, before determining the similarity between the characteristics corresponding to the multimedia material to be processed and the characteristics corresponding to the candidate editing template, determining a candidate editing template and obtaining a set of multimedia works including a preset number or more historical multimedia works corresponding to the candidate editing template; Based on the content features of the historical multimedia works in the set of the multimedia works and / or the user attributes corresponding to the historical multimedia works, determining the features corresponding to the candidate editing template; further includes.

[0009] In an alternative embodiment, determining the features corresponding to the candidate editing template based on the content features of the historical multimedia works in the set of the multimedia works and / or the user attributes corresponding to the historical multimedia works includes: determining a target tag for representing the content features of the historical multimedia works in the set of the multimedia works and / or the user attributes corresponding to the multimedia works; determining the proportion occupied by the historical multimedia works having the target tag in the set of multimedia works corresponding to the candidate editing template; determining, as a feature quantity in the dimension of the target tag of the candidate editing template, the ratio between the proportion and the average proportion corresponding to the target tag, where the feature quantity is used to represent the degree of applicability of the candidate editing template from the dimension of the target tag, and the average proportion is used to represent the average value of the proportion occupied by the historical multimedia works having the target tag in the set of multimedia works corresponding to each template in a preset template recommendation pool; constructing a template feature vector corresponding to the candidate editing template based on the feature quantity in the dimension of the target tag of the candidate editing template; includes.

[0010] In an alternative embodiment, before constructing the template feature vector corresponding to the candidate editing template based on the feature quantity in the dimension of the target tag of the candidate editing template, Determining the distribution stability of the feature amounts in the dimension of the target tag of each template in the preset template recommendation pool, wherein the distribution stability is used to represent the degree of difference in the dimension of the target tag of each template in the preset template recommendation pool, Determining the first K target tags with relatively low distribution stability as the recommended standard tags, where K is an integer greater than zero, further includes, Based on the feature amounts in the dimension of the target tag of the candidate editing template, constructing a template feature vector corresponding to the candidate editing template is, including constructing a template feature vector corresponding to the candidate editing template based on the feature amounts corresponding to the candidate editing template in the recommended standard tags.

[0011] In an alternative embodiment, obtaining the features of the multimedia material to be processed is, Based on the tags of the multimedia material to be processed, determining the feature amounts in the dimension of the recommended standard tags of the multimedia material to be processed, Based on the feature amounts in the dimension of the recommended standard tags of the multimedia material to be processed, determining a feature vector corresponding to the multimedia material to be processed, including. Determining the similarity between the features corresponding to the multimedia material to be processed and the features corresponding to the candidate editing template is, including calculating the similarity between the feature vector corresponding to the multimedia material to be processed and the template feature vector corresponding to the candidate editing template.

[0012] In an alternative embodiment, before determining the similarity between the features corresponding to the multimedia material to be processed and the features corresponding to the candidate editing template, Further comprising selecting a template in a preset template recommendation pool based on the attribute information of the multimedia material to be processed, including the material type, the number of materials, the material duration, and / or the material screen ratio, to obtain a candidate editing template.

[0013] In one selectable embodiment, the method is applied to a client. When it is determined that the similarity meets a preset matching condition, after determining the candidate editing template as the target editing template corresponding to the multimedia material to be processed, Further comprising editing the multimedia material to be processed according to the editing method indicated by the target editing template to obtain a target multimedia work.

[0014] In one selectable embodiment, the target editing template includes a plurality of candidate editing templates. Before editing the multimedia material to be processed according to the editing method indicated by the target editing template to obtain a target multimedia work, Further comprising displaying the plurality of candidate editing templates. Editing the multimedia material to be processed according to the editing method indicated by the target editing template to obtain a target multimedia work means In response to a selection operation for any one of the plurality of candidate editing templates, editing the multimedia material to be processed according to the editing method indicated by the template corresponding to the selection operation to obtain a target multimedia work.

[0015] In a second aspect, the present disclosure further A feature determination module for obtaining the features of the multimedia material to be processed, including the content features of the multimedia material to be processed and / or the user attributes corresponding to the multimedia material to be processed, A similarity determination module for determining the similarity between the features corresponding to the multimedia material to be processed and the features corresponding to the candidate editing template determined based on the historical multimedia works that are multimedia works edited in the editing method indicated by the candidate editing template, A template determination module for determining the candidate editing template as a target editing template corresponding to the multimedia material to be processed when it is determined that the similarity satisfies a preset matching condition, where the target editing template is used to instruct that the multimedia material to be processed be edited into a target multimedia work according to the editing method. Provided is a template recommendation device including the above.

[0016] In a third aspect, the present disclosure provides a computer-readable storage medium storing instructions for causing the terminal device to implement the above method when executed on the terminal device.

[0017] In a fourth aspect, the present disclosure provides an electronic device including a processor and a memory for storing instructions executable by the processor, where the processor reads the executable instructions from the memory and executes the instructions to implement the above method.

[0018] In a fifth aspect, the present disclosure provides a computer program product including a computer program / instruction for implementing the above method when executed by a processor.

[0019] The template recommendation method according to an embodiment of the present disclosure determines a target editing template corresponding to a multimedia material to be processed by calculating a similarity between a feature corresponding to the multimedia material to be processed and a feature corresponding to a candidate editing template. Since the similarity between the feature of the multimedia material to be processed and the feature of the candidate editing template can represent the degree of matching between the multimedia material to be processed and the candidate editing template, by determining a candidate editing template that satisfies a preset matching condition as the target editing template, the degree of matching between the target editing template and the multimedia material to be processed can be improved, and the user satisfaction with the template recommendation can be improved.

Brief Description of the Drawings

[0020] Referring to the following specific embodiments in conjunction with the drawings, the above and other features, advantages, and aspects of each embodiment of the present disclosure will become more apparent. Throughout the drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and the components and elements are not necessarily drawn to scale.

[0021]

Figure 1

[0022]

Figure 2

[0023]

Figure 3

[0024]

Figure 4

[0025]

Figure 5

[0026]

Figure 6

Embodiments for Carrying Out the Invention

[0027] Hereinafter, with reference to the drawings, embodiments of the present disclosure will be described in more detail. Although some embodiments of the present disclosure are shown in the drawings, the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments described herein. Rather, it should be understood that these embodiments are provided to enable a clearer and more complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are used only for illustration and not for limiting the scope of protection of the present disclosure.

[0028] It should be understood that each step described in the method embodiments of the present disclosure may be executed in a different order and / or in parallel. Furthermore, the method embodiments may include the addition of steps and / or the omission of the execution of the steps shown. The scope of the present disclosure is not limited in this regard.

[0029] As used herein, the term "comprising" and its variations are open inclusion, that is, "including but not limited to". The term "based on" means "at least partially based on". The term "one embodiment" means "at least one embodiment", the term "another embodiment" means "at least one another embodiment", and the term "some embodiments" means "at least some embodiments". Related definitions of other terms are given in the following description.

[0030] Note that concepts such as "first" and "second" mentioned in this disclosure are only used to distinguish different devices, modules, or units, and do not limit the order or interdependence of the functions executed by these devices, modules, or units.

[0031] Note that the modifiers "one" and "a plurality" mentioned in this disclosure are not restrictive but exemplary, and those skilled in the art should understand that they should be understood as "one or more" unless otherwise explicitly stated in the context.

[0032] The names of messages or information that interact between multiple devices in the embodiments of this disclosure are only used for explanation and are not used to limit the scope of these messages or information.

[0033] FIG. 1 is a schematic diagram of the flow of a template recommendation method according to an embodiment of this disclosure. The method can be executed by a template recommendation device. The device can be realized by software and / or hardware and can usually be incorporated into an electronic device.

[0034] As shown in FIG. 1, the template recommendation method can include the following steps.

[0035] In step 101, obtain the characteristics of the multimedia material to be processed, including the content characteristics of the multimedia material to be processed and / or the user attributes corresponding to the multimedia material to be processed.

[0036] The multimedia material to be processed means the multimedia material for which there is a processing request. In different application scenarios, the number of multimedia materials to be processed may be different, and the type of multimedia materials to be processed may also be different. The type of multimedia material to be processed can include any one or more of videos, images, and audios, but is not limited thereto.

[0037] In one possible embodiment, the multimedia material to be processed can include one or more types of materials among videos, images, and audios, and the number of materials of each type is not limited.

[0038] Since the material content of the multimedia material to be processed itself can represent the characteristics of the multimedia material to be processed, based on the material content of the multimedia material to be processed, the content characteristics of the multimedia material to be processed can be extracted as the characteristics of the multimedia material to be processed. The content characteristics may be characteristics representing the multimedia material to be processed from the dimension of the content of the multimedia material to be processed. Specifically, the content characteristics can be obtained by content analysis of the multimedia material to be processed. The content characteristics include, but are not limited to, any one or more of landscapes, gourmet foods, sports, and babies.

[0039] It should be understood that the multimedia material to be processed may include a portrait. Therefore, the content characteristics may include the character attribute characteristics in the multimedia material to be processed. The character attribute characteristics may be obtained by methods such as face attribute analysis. This embodiment is not limited to the character attribute characteristics. For example, any of age group, gender, expression, number of people, etc. may also be regarded as character attribute characteristics. For example, when it is analyzed that two people are included in the multimedia material to be processed by performing face attribute analysis on the multimedia material to be processed, the characteristic of "two people" can be added to the content characteristics. Also, when it is analyzed that a baby is included in the multimedia material to be processed, the characteristic of "baby" can be added to the content characteristics.

[0040] In addition, since users who select multimedia materials to be processed often expect to be able to recommend templates that suit their preferences, in the embodiments of the present disclosure, in order to construct the user attributes of the multimedia materials to be processed, the characteristics of the user who selects the multimedia materials to be processed are also regarded as the characteristics of the multimedia materials to be processed. Specifically, the user attributes are the characteristics representing the multimedia materials to be processed from the user attribute dimensions of the user who selects the multimedia materials to be processed. Specifically, they can be obtained by analyzing the user who selects the multimedia materials to be processed. The user attributes include, but are not limited to, any one or more of the user's age, gender, occupation, and preferences. For example, if the user who selects the multimedia materials to be processed is female, the user attribute of "female" can be added to the multimedia materials to be processed.

[0041] It should be understood that the characteristics of the multimedia materials to be processed can be related to the act of the user selecting the multimedia materials to be processed. For example, during a certain festival, when a user selects multimedia materials to be processed for generating a multimedia work, generally, an editing template is recommended to the user according to the characteristics of the festival. Therefore, in the embodiments of the present disclosure, in order to recommend an editing template to the user, the selected time, etc. of the multimedia materials to be processed can also be regarded as the characteristics of the multimedia materials to be processed.

[0042] In one selectable embodiment, the characteristics of the multimedia materials to be processed can be represented by corresponding feature vectors. The method for determining the feature vectors will be introduced in subsequent embodiments, so the description is omitted here.

[0043] In step 102, determine the similarity between the features corresponding to the multimedia material to be processed and the features corresponding to the candidate editing template, which are determined based on the historical multimedia works that are multimedia works edited in the editing method indicated by the candidate editing template.

[0044] The candidate editing template refers to an editing template that can be selected to process the multimedia material to be processed. Generally, each candidate editing template corresponds to one or more historical multimedia works to which the template is applied.

[0045] It should be understood that historical multimedia works to which the same candidate editing template is applied generally include similar features. Therefore, the features corresponding to the candidate editing template can be determined based on the historical multimedia works to which the same candidate editing template is applied. The features can be used to represent the features of multimedia works suitable for applying the corresponding candidate editing template. The features of the candidate editing template can also be represented from the perspective of the editing template application scenario.

[0046] Similar to the features of the multimedia material to be processed, in one selectable embodiment, the features corresponding to the candidate editing template can be represented by the corresponding feature vector. The method for determining the feature vector will be introduced in subsequent embodiments, so the description is omitted here.

[0047] In the embodiments of the present disclosure, in order to obtain a candidate editing template with a relatively high degree of matching with the multimedia material to be processed, recommend an editing template for the multimedia material to be processed based on the feature similarity between the multimedia material to be processed and the candidate editing template. The method for determining the similarity can be selected according to the specific application scenario, and this embodiment is not limited.

[0048] In step 103, when it is determined that the similarity meets the preset matching condition, the candidate editing template is determined as the target editing template corresponding to the multimedia material to be processed. The target editing template is used to instruct the multimedia material to be processed to be edited into the target multimedia work according to the editing method.

[0049] It should be understood that the larger the similarity value, the higher the degree of matching between the corresponding candidate editing template and the multimedia material to be processed. Conversely, the smaller the similarity value, the lower the degree of matching between the corresponding candidate editing template and the multimedia material to be processed. Therefore, in the embodiments of the present disclosure, the matching condition for screening the candidate editing template is preset, and based on whether the similarity between the features corresponding to the multimedia material to be processed and the features corresponding to the candidate editing template meets the preset matching condition, the target editing template corresponding to the multimedia material to be processed can be determined. The preset matching condition may be greater than a preset similarity threshold or the like. The number of the target editing templates may be one or more. Each target editing template corresponds to the associated editing method, and there are multiple types of operations that can be performed on the multimedia material to be processed by the editing method, including any one or more of adding filters, switching scenes, and adding music, but is not limited thereto. The target editing template is used to instruct the multimedia material to be processed to be edited into the target multimedia work according to the preset editing method.

[0050] Note that there are multiple types of methods for determining the target editing template, which can be selected according to the application scenario, but the present embodiment is not limited thereto. For example, there are the following two methods.

[0051] As Method 1, in the embodiments of the present disclosure, each calculated similarity can be sorted in descending order, and the first N candidate editing templates can be used as the target editing template corresponding to the multimedia material to be processed.

[0052] As Method 2, in the embodiments of the present disclosure, each calculated similarity can be sorted in descending order, and N can be randomly selected from the first M candidate editing templates to be used as the target editing template. Generally, the styles of several candidate editing templates with the highest similarity values are likely to be similar. This method can enrich the styles of the target editing templates presented to the user and improve the user experience.

[0053] The template recommendation method according to the embodiments of the present disclosure determines the target editing template corresponding to the multimedia material to be processed by calculating the similarity between the features corresponding to the multimedia material to be processed and the features corresponding to the candidate editing templates. Since the similarity between the features of the multimedia material to be processed and the features of the candidate editing templates can represent the degree of matching between the multimedia material to be processed and the candidate editing templates, by determining the candidate editing template that meets the preset matching condition as the target editing template, the degree of matching between the target editing template and the multimedia material to be processed can be improved, and the user satisfaction with the template recommendation can be improved.

[0054] In addition, the embodiments of the present disclosure can recommend different target editing templates for multimedia materials to be processed with different features, avoid recommending a single popular template to all users, and make the template recommendation more diverse.

[0055] Based on the above embodiments, before determining the similarity between the features corresponding to the multimedia material to be processed and the features corresponding to the candidate editing templates, first determine the features corresponding to the candidate editing templates.

[0056] To more clearly explain the method for determining the features corresponding to the candidate editing template, refer to FIG. 2, which is a schematic diagram of the flow of the method for determining the features of the candidate editing template according to an embodiment of the present disclosure. FIG. 2 includes the following steps.

[0057] In step 201, determine the candidate editing template and obtain a set of multimedia works that includes more historical multimedia works than a preset number corresponding to the candidate editing template.

[0058] Note that the target editing template recommended to the user can be determined from a preset template recommendation pool based on the similarity to the multimedia material to be processed. In order to improve the quality of the target editing template recommended to the user, in the embodiments of the present disclosure, the templates in the preset template recommendation pool can be screened to determine the candidate editing template in the embodiments of the present disclosure.

[0059] Note that the set of multimedia works includes historical multimedia works. The historical multimedia works in the same set of multimedia works are those processed using the same candidate editing template.

[0060] There are multiple methods for screening the templates in the preset template recommendation pool to determine the candidate editing template, and they can be selected according to specific application scenarios. However, this embodiment is not limited. For example, there are the following two methods.

[0061] As Method 1, in the embodiments of the present disclosure, templates can be screened according to the number of historical multimedia works corresponding to the templates in a preset template recommendation pool. Specifically, an integer Z can be preset. The value of Z can be set according to specific application scenarios. Obtain the set of multimedia works corresponding to the template. In the set of multimedia works, if the number of historical multimedia works is greater than Z, it is considered that the template has passed the screening, and the templates that have passed the screening can be used as candidate editing templates.

[0062] As Method 2, in the embodiments of the present disclosure, templates can be screened according to the number of historical multimedia works corresponding to the templates in a preset template recommendation pool and the number of playback times of each historical multimedia work. Specifically, an integer Y and an integer Z can be preset. The values of Z and Y can be set according to specific application scenarios. Obtain the set of multimedia works corresponding to the template, and determine the number of historical multimedia works that have been played more than Y in the set. If the number is greater than Z, it is considered that the template has passed the screening, and the templates that have passed the screening can be used as candidate editing templates.

[0063] To avoid the singleness of the target editing templates recommended to users, in one selectable embodiment, it should be understood that the distribution of the types of target editing templates recommended to users can be made more uniform by continuously screening the selected candidate editing templates. For example, the candidate editing templates can be continuously screened based on the template type, and X candidate editing templates can be selected from each template type as the final candidate editing templates. However, X is an integer.

[0064] In step 202, based on the content features of the historical multimedia work and / or the user attributes corresponding to the historical multimedia work in the set of multimedia works, determine the features corresponding to the candidate editing template.

[0065] It should be understood that the features of the historical multimedia work in the set of multimedia works can include the content features of the historical multimedia work and / or the user attributes corresponding to the creator of the historical multimedia work.

[0066] The content features can represent the features of the multimedia work from the dimension of the content of the material. Specifically, the content features can be obtained by content analysis of the multimedia work. The content features include, but are not limited to, any one or more of scenery, gourmet, sports, and babies.

[0067] It should be understood that the historical multimedia work may include a portrait of a person. Therefore, the content features may include the features of the person in the historical multimedia work. The features of the person may be obtained by methods such as facial attribute analysis. This embodiment is not limited to the features of the person. For example, any of the age group, gender, expression, number of people, etc. may also be regarded as the features of the person. For example, when it is analyzed that two people are included in the historical multimedia work by performing facial attribute analysis on the historical multimedia work, the feature of "two people" can be added to the content features. When it is analyzed that a baby is included in the historical multimedia work, the feature of "baby" can be added to the content features.

[0068] The user attributes corresponding to the creator of the historical multimedia work can represent the characteristics of the historical multimedia work from the dimension of the user attributes of the user who distributed the multimedia work. Specifically, the user attributes can be obtained by analyzing the user attributes of the user of the historical multimedia work. The user attributes include, but are not limited to, one or more of the age of the creator, the gender of the creator, the occupation of the creator, and the preferences of the creator.

[0069] It should be understood that the characteristics of the historical multimedia work can be related to the act of the user distributing the historical multimedia work. For example, during a certain festival, when the user distributes the historical multimedia work, generally, a template is selected according to the characteristics of the festival. Therefore, the embodiments of the present disclosure can also use the distribution time of the historical multimedia work, etc., as the characteristics of the historical multimedia work.

[0070] In addition, since the characteristics of the historical multimedia work to which the candidate editing template is applied can represent the characteristics of the candidate editing template, the characteristics of the historical multimedia work in the set of multimedia works corresponding to the candidate editing template can be used to represent the characteristics of the candidate editing template.

[0071] In the embodiments of the present disclosure, the historical multimedia work can have tags for representing the content characteristics of the historical multimedia work and / or the user attributes corresponding to the creator of the historical multimedia work. Based on the tags of the historical multimedia work in the set of multimedia works, a template feature vector for representing the characteristics of the candidate editing template corresponding to the candidate editing template can be determined. The determination method can be selected according to specific application scenarios, and this embodiment does not limit this, for example, there are the following two methods.

[0072] As Method 1, based on the tags of historical multimedia works, determine the feature vector of each historical multimedia work, and calculate the average value of the feature vectors of each historical multimedia work as the template feature vector corresponding to the candidate editing template. The specific calculation method is omitted here.

[0073] As Method 2, it can be realized based on the following Steps 1 to 4. Specifically, it includes the following steps.

[0074] In Step 1, determine the content features of the historical multimedia works in the set of multimedia works and / or the target tags for representing the user attributes corresponding to the multimedia works.

[0075] In the embodiments of the present disclosure, the template feature vector corresponding to the candidate editing template can include feature quantities in multiple dimensions. Each dimension of the template feature vector is respectively associated with a specific tag. The feature quantity of each dimension is used to represent the compatibility of the candidate editing template of the corresponding tag. The higher the feature quantity, the higher the compatibility between the multimedia material to be processed having the corresponding tag and the candidate editing template.

[0076] In actual applications, before determining the template feature vector corresponding to the candidate editing template, first determine the tags associated with each dimension of the template feature vector, and use them as target tags respectively.

[0077] In Step 2, determine the proportion of the historical multimedia works having the target tags in the set of multimedia works corresponding to the candidate editing template.

[0078] In an embodiment of the present disclosure, after obtaining a set of multimedia works corresponding to a candidate editing template, tags of each historical multimedia work in the set of multimedia works are determined. Then, based on the tags of each historical multimedia work, a ratio of the historical multimedia works having a target tag in the set of multimedia works is determined. For example, when the target tag is the "baby" tag, in the set of multimedia works corresponding to the candidate editing template, the number of historical multimedia works having the "baby" tag is determined to be 1000, while the total number of historical multimedia works in the set of multimedia works is 10000, so the ratio occupied by the historical multimedia works having the "baby" tag is 10%.

[0079] It should be understood that the ratio occupied by the historical multimedia works having a target tag can represent the saliency of the target tag for the candidate editing template. If most of the historical multimedia works to which the candidate editing template is applied have the "baby" tag, the "baby" tag will have relatively high saliency for the candidate editing template.

[0080] In step 3, a ratio between the ratio and the average ratio corresponding to the target tag is determined as a feature amount in the dimension of the target tag of the candidate editing template. Here, the feature amount is used to represent the degree of application of the candidate editing template from the dimension of the target tag. The average ratio is used to represent the average value of the ratios occupied by the historical multimedia works having the target tag in the sets of multimedia works corresponding to each template in a preset template recommendation pool.

[0081] In an embodiment of the present disclosure, a preset template recommendation pool can be constructed. Here, the template recommendation pool includes templates determined based on information such as the number of historical multimedia works corresponding to the template, the number of "likes" of the historical multimedia works, and the number of reproductions of the historical multimedia works. The average ratio corresponding to the target tag in the embodiment of the present disclosure can be used to represent the average prominence degree for all templates in the preset template recommendation pool of the target tag.

[0082] Specifically, the calculation method of the average ratio corresponding to the target tag may be to obtain, as the average ratio, the ratio of the historical multimedia works having the target tag in the set of multimedia works corresponding to all templates for the templates in the preset template recommendation pool. Alternatively, the average value of the ratio of the historical multimedia works having the target tag in the set of multimedia works corresponding to each template may be obtained as the average ratio corresponding to the target tag. Other methods for calculating the average ratio corresponding to the target tag will not be described here.

[0083] To represent the degree of applicability of the candidate editing template of the target tag, the embodiment of the present disclosure uses the ratio between the ratio of the historical multimedia works having the target tag in the set of multimedia works corresponding to the candidate editing template and the average ratio corresponding to the target tag as the feature amount in the dimension of the target tag of the candidate editing template.

[0084] For example, when the target tag is "baby", the proportion of historical multimedia works using the candidate editing template that have the "baby" tag is 10%. If the average proportion of historical multimedia works corresponding to all templates in the preset template recommendation pool that have the "baby" tag is 2%, it means that the "baby" tag has relatively high saliency for the candidate editing template. Therefore, the feature amount in the dimension of the "baby" tag of the candidate editing template is 10%÷2% = 5.

[0085] In step 4, based on the feature amount in the dimension of the target tag of the candidate editing template, a template feature vector corresponding to the candidate editing template is constructed.

[0086] It should be understood that each dimension in the template feature vector is associated with a tag respectively. After determining the feature amount corresponding to each tag, a template feature vector corresponding to the candidate editing template is constructed based on the feature amount corresponding to each tag.

[0087] In the embodiments of the present disclosure, the feature amount of the target tag is determined by the ratio between the proportion and the average proportion, and the difference between templates is more clearly shown.

[0088] In an embodiment of the present disclosure, before step 4, tags corresponding to each dimension can also be selected based on the distribution stability of feature amounts in each dimension. It should be understood that for all templates, the distribution stability of the feature amounts of the target tags can represent the degree of difference with respect to the target tags in each template. That is, based on the distribution stability of the feature amounts, tags with unclear differences among the templates can be selected and excluded. Therefore, the selected tags can more clearly characterize the differences among the templates, making it easier to realize template recommendation. At the same time, the dimension of the template feature vector can be reduced by tag selection. Therefore, the computational amount for calculating the similarity of the feature vectors corresponding to the multimedia materials to be processed is reduced, and the efficiency of template recommendation is improved as a whole.

[0089] In one selectable embodiment, first, the distribution stability of the feature amounts in the dimension of the target tags of each template in a preset template recommendation pool is determined. Then, the first K target tags with relatively low distribution stability are determined as the recommended standard tags. Here, K is an integer greater than zero. It should be understood that the lower the distribution stability of a tag, the greater the difference in feature amounts among different templates, and the difference among different templates can be reflected. Therefore, the first K target tags with relatively low distribution stability can be selected and determined as the recommended standard tags. Here, K is an integer greater than zero.

[0090] In an embodiment of the present disclosure, the distribution stability can be selected according to specific application scenarios. The distribution stability includes, but is not limited to, any one of variance, standard deviation, and mean deviation.

[0091] In one selectable embodiment, for the target tag, the variance of the feature amounts in the dimension of the target tag of each template is calculated, and based on the variance corresponding to each tag, the first K tags with large variances can be determined as the recommended standard tags. The first K tags with large variances can represent the differences between different templates.

[0092] After the recommended standard tags are determined, accordingly, in step 4 described above, based on the feature amounts of the candidate editing template in the dimension of the target tag, constructing the template feature vector corresponding to the candidate editing template is It can include constructing the template feature vector corresponding to the candidate editing template based on the feature amounts corresponding to the candidate editing template with the recommended standard tags. Specifically, using the recommended standard tags as the target tags, steps 2 to 4 above are sequentially executed to complete the construction of the template feature vector corresponding to the candidate editing template.

[0093] In the embodiments of the present disclosure, after determining the K recommended standard tags, the template feature vectors corresponding to each template are further determined, and based on each template feature vector, an N×K feature matrix is constructed. Here, N is used to represent the number of templates, and each dimension corresponding to each column in the N×K feature matrix is normalized. The method of this normalization process includes, but is not limited to, either L1 normalization or softmax.

[0094] In one selectable embodiment, the feature vector corresponding to the multimedia material to be processed is associated with each dimension of the template feature vector, the tags of the corresponding dimensions are the same, and the total number of included dimensions is also the same. For example, if the first dimension in the feature vector corresponding to the multimedia material to be processed is the feature amount of the "baby" tag, the first dimension of the template feature vector is also the feature amount of the "baby" tag.

[0095] Therefore, after the recommended standard tags are determined, the characteristics of the multimedia material to be processed can be obtained. The expression of the characteristics can be a feature vector corresponding to the multimedia material. Specifically, First, based on the tags of the multimedia material to be processed, determine the feature amount of the multimedia material to be processed in the dimension of the recommended standard tags, and then, based on the feature amount of the multimedia material to be processed in the dimension of the recommended standard tags, determine the feature vector corresponding to the multimedia material to be processed. This can include the above steps.

[0096] In the embodiments of the present disclosure, the methods for determining the feature vector corresponding to the multimedia material to be processed can include multiple types. Such methods are selected according to specific application scenarios, and the present embodiment does not limit this. The following two methods are included, but not limited thereto.

[0097] As Method 1, pre-determine the recommended standard tags corresponding to each dimension of the feature vector corresponding to the multimedia material to be processed, and then, based on whether each recommended standard tag exists in the tags of the multimedia material to be processed, determine the feature amount of the corresponding dimension. Specifically, if the recommended standard tag corresponding to a certain dimension exists in the tags of the multimedia material to be processed, determine the feature amount of that dimension as 1; otherwise, determine it as 0, and sequentially obtain the binarized feature vector corresponding to the multimedia material to be processed.

[0098] For example, when the multimedia material to be processed includes pixel materials, the correspondence between the dimension in the feature vector corresponding to the multimedia material to be processed and the recommended standard tag can be determined in advance. Suppose the feature vector includes three dimensions, and the recommended standard tags corresponding to each dimension are respectively: the first dimension corresponds to the "baby" recommended standard tag, the second dimension corresponds to the "scenery" recommended standard tag, and the third dimension corresponds to the "sports" recommended standard tag. Then, it is determined whether the recommended standard tags corresponding to the above-mentioned respective dimensions are included in the tags of the multimedia material to be processed. If the tags of the multimedia material to be processed include the "baby" and "scenery" tags but do not include the "sports" tag, the feature vector corresponding to the multimedia material to be processed is [1, 1, 0] T will be obtained.

[0099] As a second method, when the multimedia material to be processed includes a plurality of materials, the feature vector corresponding to the multimedia material to be processed can be determined according to the plurality of materials. The recommended standard tags corresponding to each dimension of the feature vector corresponding to the multimedia material to be processed are determined in advance, and for each dimension of the recommended standard tag, the ratio of the number of materials in the multimedia material to be processed that include the recommended standard tag to the total number of materials can be determined, and this ratio is the feature amount of the dimension corresponding to the recommended standard tag in the feature vector. For example, suppose there are three materials in the multimedia material to be processed, and among these three materials, if the recommended standard tag corresponding to a certain dimension exists in the tags of two materials, then the feature amount of this dimension is determined to be 2 / 3, and the feature vector corresponding to the multimedia material to be processed can be obtained according to the above method.

[0100] For example, when a plurality of pixel materials are included in the multimedia material to be processed, the correspondence between the dimensions included in the feature vector corresponding to the multimedia material to be processed and the recommended standard tags can be determined in advance. The recommended standard tags corresponding to each dimension are such that the first dimension corresponds to the "baby" recommended standard tag, the second dimension corresponds to the "scenery" recommended standard tag, and the third dimension corresponds to the "sports" recommended standard tag. In a plurality of materials, if the first material has tags "baby" and "scenery", the second material has tags "baby" and "sports", and the third material has tags "baby" and "scenery", the feature vector corresponding to the multimedia material to be processed is [1, 2 / 3, 1 / 3] T It becomes like this.

[0101] Accordingly, after the feature vector corresponding to the multimedia material to be processed and the template feature vector corresponding to the candidate editing template are determined, the similarity between the feature vector corresponding to the multimedia material to be processed and the template feature vector corresponding to the candidate editing template is calculated.

[0102] The feature vector corresponding to the multimedia material to be processed can represent the features of the multimedia material to be processed, and the template feature vector corresponding to the candidate editing template can represent the features of the candidate editing template. Therefore, the similarity can be calculated for the feature vector of the multimedia material to be processed and the feature vector of the candidate editing template, and the degree of matching between the target editing template determined according to the similarity and the multimedia material to be processed is relatively high, improving the user satisfaction with the template recommendation.

[0103] Note that there are multiple methods for calculating the similarity, and this embodiment is not limited. For example, it may be either the cosine similarity or the Euclidean distance.

[0104] Based on the above embodiments, in order to reduce the computational complexity of calculating the similarity between the features corresponding to the multimedia material to be processed and the features corresponding to the candidate editing template, the candidate editing template can be pre-screened based on the matching status of the attribute information.

[0105] Specifically, based on the attribute information of the multimedia material to be processed, the templates in the preset template recommendation pool can be screened to obtain the candidate editing template. The attribute information includes the material type, the number of materials, the material time length, and / or the material screen ratio.

[0106] In general, the multimedia material to be processed has a plurality of pieces of attribute information that can be used to determine whether a template has the ability to process the multimedia material to be processed. The attribute information includes any one or more of the material type, the number of materials, the material time length, and the material screen ratio, but is not limited thereto. The embodiments of the present disclosure can screen the templates in the preset template recommendation pool based on the attribute information of the multimedia material to be processed to obtain the candidate editing template. It should be understood that the candidate editing template can be used to process the multimedia material to be processed.

[0107] For example, when the multimedia to be processed includes a plurality of pixel materials, since the material type in the attribute information of the multimedia material to be processed is the image type, among the preset template recommendation pools, the templates that cannot process the multimedia material of the image type are screened out and excluded to obtain the candidate editing template. That is, after screening, among the attribute information, the templates that do not include the image type in the material type are screened out and excluded, but the templates that include the image type in the material type are retained and used as the candidate editing template for subsequent template recommendations.

[0108] After selecting templates in a preset template recommendation pool, embodiments of the present disclosure can calculate similarity based on the features corresponding to the candidate editing templates obtained by the selection and the features corresponding to the multimedia material to be processed, thereby reducing the amount of calculation.

[0109] In the template recommendation method according to an embodiment of the present invention, before calculating the similarity, templates in a preset template recommendation pool are selected, and templates that do not match the attribute information of the multimedia material to be processed are excluded, reducing the amount of data involved in the calculation of the similarity and improving the overall efficiency of template recommendation.

[0110] In one application scenario, the template recommendation method according to an embodiment of the present invention can be applied to a client. The client includes, but is not limited to, any one or more of a web client and a mobile client.

[0111] As shown in FIG. 3, FIG. 3 is a data interaction diagram of the template recommendation method according to an embodiment of the present disclosure.

[0112] In step 301, the client acquires the features of the multimedia material to be processed.

[0113] Before the features of the multimedia material to be processed are determined, first, the multimedia material to be processed is selected on the client by the user, and then the client extracts the features of the multimedia material to be processed. Specifically, the extraction method of the features of the multimedia material to be processed is not limited.

[0114] In step 302, the client sends a request to the server to request the server to send the features corresponding to the latest version of the candidate editing template.

[0115] In step 303, the server responds to the request sent by the client in step 302 and sends the features corresponding to the latest version of the candidate editing template to the client.

[0116] In step 304, the client determines the similarity between the features corresponding to the multimedia material to be processed and the features corresponding to the candidate editing template. If it is determined that the similarity meets the preset matching condition, the candidate editing template is determined as the target editing template corresponding to the multimedia material to be processed.

[0117] Determine the ID (Identity document) corresponding to the target editing template corresponding to the multimedia material to be processed, and use this ID as the unique identifier corresponding to the template. It should be understood that the target editing template may be one or more. Therefore, the ID corresponding to the target editing template determined by the client may be one or more.

[0118] In step 305, the client sends the ID corresponding to the target editing template to the server.

[0119] In step 306, after receiving the ID, the server acquires the target editing template corresponding to the ID.

[0120] In step 307, the server sends the target editing template to the client.

[0121] In step 308, the client generates a target multimedia work. Specifically, When the number of target editing templates is one, the client edits the multimedia material to be processed according to the editing method indicated by the target editing template to obtain a target multimedia work.

[0122] When there are multiple target editing templates, the client can display multiple candidate editing templates to the user. The user selects a target editing template according to their preferences. In response to a selection operation for any of the multiple candidate editing templates, the client edits the multimedia material to be processed according to the editing method indicated by the template corresponding to the selection operation to obtain a target multimedia work.

[0123] In an embodiment of the present disclosure, the client enables further independent selection by the user by displaying the target editing template to the user, improves the interactivity of template recommendation, and improves the user experience.

[0124] In another application scenario, the template recommendation method according to the embodiment of the present disclosure can also be applied to a server.

[0125] As shown in FIG. 4, FIG. 4 is a data interaction diagram of another template recommendation method according to an embodiment of the present disclosure.

[0126] In step 401, the client determines the characteristics of the multimedia material to be processed. The characteristics of the multimedia material to be processed can include the feature vector of the multimedia material to be processed, or can include the tags of the multimedia material to be processed.

[0127] In actual application, before obtaining the characteristics of the multimedia material to be processed, first, the multimedia material to be processed is selected on the client by the user.

[0128] In step 402, the client sends the characteristics of the multimedia material to be processed to the server.

[0129] In step 403, the server obtains the characteristics of the multimedia material to be processed from the client, and determines the similarity between the characteristics of the multimedia material to be processed and the characteristics of the candidate editing template. If it is determined that the similarity meets the preset matching condition, the corresponding candidate editing template is determined as the target editing template corresponding to the multimedia material to be processed.

[0130] In step 404, the server sends the target editing template to the client.

[0131] In step 405, the client generates a target multimedia work. Specifically, If the number of target recommendation templates is one, the client edits the multimedia material to be processed according to the editing method indicated by the target editing template to obtain a target multimedia work.

[0132] If the number of target recommendation templates is multiple, each target recommendation template can be displayed to the user by the client. The user selects a target recommendation template according to their preference. In response to the selection operation for any of the target recommendation templates, the client generates a target multimedia work based on the selected template and the multimedia material to be processed.

[0133] The client generates a target multimedia work based on the selected template and the multimedia material to be processed. Specifically, the client can edit the multimedia material to be processed according to the editing method indicated by the template corresponding to the selection operation to obtain a target multimedia work.

[0134] The user does not need to upload the multimedia material to be processed to the server, but only needs to upload the characteristics of the multimedia material to be processed, such as feature vectors and tags. Therefore, it should be understood that leakage of the user's privacy data can be avoided.

[0135] Based on the above method embodiments, the present disclosure further provides a template recommendation device. As shown in FIG. 5, it is a schematic configuration diagram of a template recommendation device according to an embodiment of the present disclosure. The device can be implemented by software and / or hardware and can usually be incorporated into an electronic device. As shown in FIG. 5, the template recommendation device 500 includes a feature determination module 501 for obtaining the characteristics of the multimedia material to be processed, including the content characteristics of the multimedia material to be processed and / or the user attributes corresponding to the multimedia material to be processed; a similarity determination module 502 for determining the similarity between the characteristics corresponding to the multimedia material to be processed and the characteristics corresponding to the candidate editing template, which are determined based on a historical multimedia work that is a multimedia work edited in an editing method indicated by the candidate editing template; a template determination module 503 for determining the candidate editing template as a target editing template corresponding to the multimedia material to be processed when it is determined that the similarity meets a preset matching condition. The target editing template is used to instruct that the multimedia material to be processed be edited into a target multimedia work according to the editing method. and can include.

[0136] The template recommendation device includes an acquisition module for determining a candidate editing template and obtaining a set of multimedia works including a number of historical multimedia works more than a preset number corresponding to the candidate editing template; A first determination module for determining features corresponding to the candidate editing template based on content features of historical multimedia works in the set of multimedia works and / or user attributes corresponding to the historical multimedia works; It may further include.

[0137] The first determination module Determining a target tag for representing content features of historical multimedia works in the set of multimedia works and / or user attributes corresponding to the multimedia works; Determining the proportion of historical multimedia works having the target tag in the set of multimedia works corresponding to the candidate editing template; Determining, as a feature quantity of the candidate editing template in the dimension of the target tag, a ratio between the proportion and an average proportion corresponding to the target tag, where the feature quantity is used to represent the degree of applicability of the candidate editing template from the dimension of the target tag, and the average proportion is used to represent an average value of the proportion of historical multimedia works having the target tag in the set of multimedia works corresponding to each template in a preset template recommendation pool; Constructing a template feature vector corresponding to the candidate editing template based on the feature quantity of the candidate editing template in the dimension of the target tag; It may be further used for.

[0138] The first determination module Determining the distribution stability of the feature quantities of each template in the preset template recommendation pool in the dimension of the target tag, where the distribution stability is used to represent the degree of difference of each template in the preset template recommendation pool in the dimension of the target tag; Determining the first K target tags with relatively low distribution stability as recommended standard tags, where K is an integer greater than zero, may be further used for this purpose. Based on the feature amount of the candidate editing template in the dimension of the target tag, constructing a template feature vector corresponding to the candidate editing template, may include constructing a template feature vector corresponding to the candidate editing template based on the corresponding feature amount of the candidate editing template in the recommended standard tag.

[0139] The first determination module determines the feature amount of the multimedia material to be processed in the dimension of the recommended standard tag based on the tag of the multimedia material to be processed, and may further be used to determine a feature vector corresponding to the multimedia material to be processed based on the feature amount of the multimedia material to be processed in the dimension of the recommended standard tag. Determining the similarity between the feature of the multimedia material to be processed and the feature corresponding to the candidate editing template may include calculating the similarity between the feature vector corresponding to the multimedia material to be processed and the template feature vector corresponding to the candidate editing template.

[0140] The template recommendation device may further include a selection module for obtaining candidate editing templates by selecting templates in a preset template recommendation pool based on the attribute information of the multimedia material to be processed, including material type, number of materials, material duration, and / or material screen ratio.

[0141] The device is applied to a client, and the template recommendation device It may further include a generation module for editing the multimedia material to be processed according to the editing method indicated by the target editing template to obtain a target multimedia work.

[0142] The template recommendation device may further include a display module for displaying the plurality of candidate editing templates. The generation module is used to edit the multimedia material to be processed according to the editing method indicated by the template corresponding to the selection operation in response to a selection operation for any one of the plurality of candidate editing templates to obtain a target multimedia work. The template recommendation device according to an embodiment of the present disclosure can execute the template recommendation method according to any embodiment of the present disclosure, and has a functional module and beneficial effects corresponding to the execution method.

[0143] In addition to the above methods and devices, an embodiment of the present disclosure further provides a computer-readable storage medium storing instructions for realizing the template recommendation method according to an embodiment of the present disclosure when executed on a terminal device.

[0144] An embodiment of the present disclosure further provides a computer program product including a computer program / instructions for realizing the template recommendation method according to any embodiment of the present disclosure when executed by a processor.

[0145] FIG. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present disclosure.

[0146] Hereinafter, specifically refer to FIG. 6 showing a configuration schematic diagram suitable for use in realizing the electronic device 600 according to an embodiment of the present disclosure. The electronic device 600 according to an embodiment of the present disclosure can include portable terminals such as mobile phones, notebook computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (tablets), PMPs (Portable Multimedia Players), in-vehicle terminals (for example, car navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc., but is not limited thereto. The electronic device shown in FIG. 6 is merely an example and does not limit the functions and usage ranges of the embodiments of the present disclosure.

[0147] As shown in FIG. 6, the electronic device 600 can include a processing device (for example, a central processor, a graphics processor, etc.) 601 that can execute various appropriate operations and processes according to a program stored in a read-only memory (ROM) 602 or a program loaded from a storage device 608 into a random access memory (RAM) 603. Various programs and data necessary for the operation of the electronic device 600 are also stored in the RAM 603. The processing device 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0148] Generally, an input device 606 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc., an output device 607 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc., a storage device 608 including, for example, a magnetic tape, a hard disk, etc., and a communication device 609 can be connected to an I / O interface 605. The communication device 609 can allow the electronic device 600 to communicate wirelessly or wiredly with other devices for data exchange. FIG. 6 shows an electronic device 600 having various devices, but it should be understood that it is not necessarily required to implement or include all the shown devices. More or fewer devices can be alternatively implemented or included.

[0149] In particular, according to an embodiment of the present disclosure, the process described above with reference to the flowchart can be realized as a computer software program. For example, an embodiment of the present disclosure includes a computer program product including a computer program mounted on a non-transitory computer-readable medium. This computer program includes program code for executing the method shown in the flowchart. In such an embodiment, the computer program may be downloaded and installed from a network by the communication device 609, or installed from the storage device 608, or installed from the ROM 602. When the computer program is executed by the processing device 601, the above functions limited to the template recommendation method according to the embodiment of the present disclosure are executed.

[0150] Note that the computer-readable medium according to the present disclosure may be a computer-readable signal medium, a computer-readable storage medium, or any combination of the two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of the computer-readable storage medium may include, but are not limited to, electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical memory devices, magnetic memory devices, or any suitable combination of the above. In the present disclosure, the computer-readable storage medium may be any tangible medium that includes or stores a program that can be used by or in combination with an instruction execution system, apparatus, or device. In the present disclosure, the computer-readable signal medium may include a data signal propagated in a baseband or propagated as part of a carrier wave on which a computer-readable program code is carried. Such a propagated data signal may take various forms, including, but not limited to, electromagnetic signals, optical signals, or any suitable combination of the above. Also, the computer-readable signal medium may be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transmit a program for use by or in combination with an instruction execution system, apparatus, or device. The program code included in the computer-readable medium can be transmitted through any suitable medium, including, but not limited to, wires, optical cables, RF (radio frequency), or any suitable combination of the above. In some embodiments, the client and the server can communicate using any currently known or future-researched / developed network protocol such as HTTP (HyperText Transfer Protocol), and can interconnect with digital data communication of any form or medium (e.g., communication network). Examples of communication networks include local area networks ("LAN"), wide area networks ("WAN"), extranets (e.g., the Internet), end-to-end networks (e.g., ad hoc end-to-end networks), and networks currently known or future-researched / developed, etc.

[0151] The computer-readable medium may be included in the electronic device or may exist separately without being incorporated into the electronic device.

[0152] The computer-readable medium stores one or more programs, and when the one or more programs are executed by the electronic device, the electronic device is caused to obtain characteristics of the multimedia material to be processed, including content characteristics of the multimedia material to be processed and / or user attributes corresponding to the multimedia material to be processed, and to determine a similarity between the characteristics corresponding to the multimedia material to be processed and the characteristics corresponding to the candidate editing template, based on a historical multimedia work that is a multimedia work edited in an editing method indicated by the candidate editing template. When it is determined that the similarity satisfies a preset matching condition, the candidate editing template is determined as a target editing template corresponding to the multimedia material to be processed, and the target editing template is used to instruct that the multimedia material to be processed be edited into a target multimedia work according to the editing method. In an embodiment of the present invention, the target editing template is determined according to the similarity between the characteristics corresponding to the multimedia material to be processed and the characteristics corresponding to the candidate editing template, and the degree of matching with the multimedia material to be processed is relatively high, thus improving the user satisfaction with the template recommendation.

[0153] The computer program code for performing the operations of the present disclosure may be written in one or more programming languages, or combinations thereof. The programming languages include, but are not limited to, object-oriented programming languages such as Java, Smalltalk, C++, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code may be executed entirely on the user computer, partially on the user computer, executed as an independent software package, partially on the user computer and partially on a remote computer, or entirely on a remote computer or server. When a remote computer is involved, the remote computer can be connected to the user computer via any type of network including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (for example, connected via the Internet using an Internet service provider).

[0154] The flowcharts and block diagrams in the drawings illustrate the architecture, functionality, and operation of systems, methods, and computer program products that can be implemented according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagram can represent one module, program segment, or part of code that includes one or more executable instructions for implementing a given logical function. It should also be noted that in some alternative implementations, the functions represented within the blocks may occur in a different order than shown in the drawings. For example, two blocks shown consecutively may actually be executed substantially in parallel, or may be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, as well as combinations of blocks in the block diagrams and / or flowcharts, may be implemented by a dedicated hardware-based system that performs a given function or operation, or may be implemented by a combination of dedicated hardware and computer instructions.

[0155] The units described in connection with the embodiments of the present disclosure may be implemented in software or in hardware. The names of the units do not limit the units themselves.

[0156] The functions described above in this specification can be executed, at least in part, by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include field programmable gate arrays (FPGA), application specific integrated circuits (ASIC), application specific standard products (ASSP), system on a chip (SOC), complex programmable logic devices (CPLD), and the like.

[0157] In the context of the present disclosure, a machine-readable medium may be a tangible medium that can contain or store a program for use by or in combination with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. The machine-readable medium can include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination thereof. More specific examples of the machine-readable storage medium include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical memory devices, magnetic memory devices, or any suitable combination thereof.

[0158] The above description is only for the better embodiments of the present disclosure and the applicable technical principles. Those skilled in the art will understand that the scope of the disclosure according to the present disclosure is not limited to the technical solutions consisting of specific combinations of the above technical features, and at the same time, without departing from the idea of the above disclosure, other technical solutions consisting of any combination of the above technical features or their equivalent features should also be included. For example, a technical solution formed by mutually replacing the above features with technical features having similar functions to those disclosed in the present disclosure (but not limited thereto).

[0159] Also, although the operations are depicted in a particular order, it should not be understood that these operations are required to be performed in the particular order shown or sequentially. In some environments, multitasking or parallel processing may be advantageous. Similarly, although some specific implementation details are included in the above discussion, these should not be construed as limiting the scope of the disclosure. Some features that are described in the context of individual embodiments may be implemented in combination in a single embodiment. Conversely, various features that are described in the context of a single embodiment may be implemented separately or in any suitable sub-combination in multiple embodiments.

[0160] The subject matter has been described in language specific to structural features and / or methodological logical acts, but it is understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described above. Conversely, the specific features and acts described above are merely exemplary forms of implementing the claims.

Claims

1. A template recommendation method, comprising: obtaining features of the multimedia material to be processed, including at least one of content features of the multimedia material to be processed and user attributes corresponding to the multimedia material to be processed; selecting templates in a preset template recommendation pool based on attribute information of the multimedia material to be processed, including at least one of material type, number of materials, material duration, and material screen ratio, to obtain candidate editing templates; determining a similarity between the features corresponding to the multimedia material to be processed and the features corresponding to the candidate editing template, based on the features corresponding to the multimedia material to be processed and a historical multimedia work which is a multimedia work edited by an editing method indicated by the candidate editing template; when it is determined that the similarity meets a preset matching condition, determining the candidate editing template as a target editing template corresponding to the multimedia material to be processed, wherein the target editing template is used to instruct that the multimedia material to be processed is edited into a target multimedia work according to the editing method; A method comprising the above.

2. Before determining the similarity between the features corresponding to the multimedia material to be processed and the features corresponding to the candidate editing template, determining a candidate editing template, and obtaining a set of multimedia works including a preset number or more of historical multimedia works corresponding to the candidate editing template; determining the features corresponding to the candidate editing template based on at least one of content features of the historical multimedia works in the set of multimedia works and user attributes corresponding to the historical multimedia works; The method according to claim 1, further comprising the above.

3. A template recommendation method, comprising: obtaining features of the multimedia material to be processed, including at least one of content features of the multimedia material to be processed and user attributes corresponding to the multimedia material to be processed; Determining the similarity between the features corresponding to the multimedia material to be processed and the features corresponding to the candidate editing template, based on a historical multimedia work that is a multimedia work edited in an editing method indicated by the candidate editing template; When it is determined that the similarity satisfies a preset matching condition, determining the candidate editing template as a target editing template corresponding to the multimedia material to be processed, wherein the target editing template is used to instruct the multimedia material to be processed to be edited into a target multimedia work according to the editing method; including; Before determining the similarity between the features corresponding to the multimedia material to be processed and the features corresponding to the candidate editing template, determining a candidate editing template and obtaining a set of multimedia works including a preset number or more historical multimedia works corresponding to the candidate editing template; determining the corresponding features of the candidate editing template based on at least one of the content features of the historical multimedia works in the set of multimedia works and the user attributes corresponding to the historical multimedia works; further including; Determining the corresponding features of the candidate editing template based on at least one of the content features of the historical multimedia works in the set of multimedia works and the user attributes corresponding to the historical multimedia works means determining a target tag for representing at least one of the content features of the historical multimedia works in the set of multimedia works and the user attributes corresponding to the multimedia works; determining the proportion of the historical multimedia works having the target tag in the set of multimedia works corresponding to the candidate editing template; Determining the ratio of the ratio between the said ratio and the average ratio corresponding to the said target tag as a feature amount of the said candidate editing template in the dimension of the said target tag, where the said feature amount is used to represent the degree of applicability of the said candidate editing template from the dimension of the said target tag, and the said average ratio is used to represent the average value of the ratio occupied by the historical multimedia works having the said target tag in the set of multimedia works corresponding to each template in a preset template recommendation pool. Constructing a template feature vector corresponding to the said candidate editing template based on the feature amount of the said candidate editing template in the dimension of the said target tag. A method including the above.

4. Before constructing a template feature vector corresponding to the said candidate editing template based on the feature amount of the said candidate editing template in the dimension of the said target tag, Determining the distribution stability of the feature amounts of each template in the preset template recommendation pool in the dimension of the said target tag, where the said distribution stability is used to represent the degree of difference of each template in the preset template recommendation pool in the dimension of the said target tag. Determining the first K target tags with relatively low said distribution stability as recommended standard tags, provided that K is an integer greater than zero. Further including the above. Constructing a template feature vector corresponding to the said candidate editing template based on the feature amount of the said candidate editing template in the dimension of the said target tag includes Constructing a template feature vector corresponding to the said candidate editing template based on the feature amount corresponding to the said candidate editing template in the said recommended standard tag. The method according to claim 3.

5. Obtaining the features of the multimedia material to be processed includes Determining the feature amount of the multimedia material to be processed in the dimension of the said recommended standard tag based on the tag of the multimedia material to be processed. Determining the corresponding feature vector of the multimedia material to be processed based on the feature amount of the multimedia material to be processed in the dimension of the said recommended standard tag. Including the above. Determining the similarity between the respective features of the multimedia material to be processed and the corresponding features of the candidate editing template includes calculating the similarity between the respective feature vectors of the multimedia material to be processed and the corresponding template feature vectors of the candidate editing template The method according to claim 4.

6. The method is applied to a client. When it is determined that the similarity satisfies a preset matching condition, after determining the candidate editing template as the target editing template corresponding to the multimedia material to be processed, further comprising editing the multimedia material to be processed according to the editing method indicated by the target editing template to obtain a target multimedia work The method according to claim 1.

7. The target editing template includes a plurality of candidate editing templates. Before editing the multimedia material to be processed according to the editing method indicated by the target editing template to obtain a target multimedia work, further comprising displaying the plurality of candidate editing templates Editing the multimedia material to be processed according to the editing method indicated by the target editing template to obtain a target multimedia work includes editing the multimedia material to be processed according to the editing method indicated by the template corresponding to the selection operation in response to a selection operation for any one of the plurality of candidate editing templates to obtain a target multimedia work The method according to claim 6.

8. A feature determination module for obtaining the features of the multimedia material to be processed, including at least one of the content features of the multimedia material to be processed and the user attributes corresponding to the multimedia material to be processed, A screening module for screening templates in a preset template recommendation pool based on the attribute information of the multimedia material to be processed, including at least one of the material type, the number of materials, the material duration, and the material screen ratio, to obtain candidate editing templates A similarity determination module for determining the similarity between the features of each of the multimedia materials to be processed and the corresponding features of the candidate editing template determined based on a historical multimedia work that is a multimedia work edited in an editing method indicated by the candidate editing template, A template determination module for determining the candidate editing template as a target editing template corresponding to the multimedia material to be processed when it is determined that the similarity satisfies a preset matching condition, where the target editing template is used to instruct that the multimedia material to be processed be edited into a target multimedia work according to the editing method, A template recommendation device comprising the above.

9. A processor, A memory for storing instructions executable by the processor, an electronic device comprising: The processor reads the executable instructions from the memory and executes the instructions to implement the method according to any one of claims 1 to 7 An electronic device.

10. A computer-readable storage medium storing instructions for causing the terminal device to implement the method according to any one of claims 1 to 7 when executed on the terminal device A computer-readable storage medium.

11. A computer program that, when executed by a processor, implements the method according to any one of claims 1 to 7

Images